TWI765839B - Parts Mounting System - Google Patents

Abstract

The plurality of inspection devices (12) on the component mounting line (10) include an inspection imaging unit (122) for acquiring inspection images (GI), a first inspection and determination unit (1233), a second inspection and determination unit (1234), and a first inspection and determination unit (1234) 3. The inspection and determination unit (1235). The first inspection determination unit (1233) determines whether the processing state of the corresponding processing device (11) is good by detecting the feature value of the inspection image (GI), and outputs the first inspection determination information (DI1) indicating the determination result. The second inspection determination unit (1234) classifies the inspection image (GI) into a first category with good processing status and a second category with poor processing status, and determines whether the processing status is good or not according to the category to which the inspection image (GI) belongs, The second inspection judgment information (DI2) indicating the judgment result is output. The third inspection determination unit (1235) outputs third inspection determination information (DI3) indicating the final determination result of whether the processing state is good or not based on the first inspection determination information (DI1) and the second inspection determination information (DI2).

Description

Parts Mounting System

The present invention relates to a component mounting system including a component mounting line including a component mounting device.

Conventionally, there has been known a component mounting system including a plurality of processing apparatuses including a component mounting apparatus for producing a component mounting substrate, and a component mounting line in which an inspection apparatus is arranged in a line. In the component mounting line, an image of the processing state of the processing device is acquired, and it is determined whether the processing state is good or not based on the image.

Patent Document 1 discloses a technique for analyzing the cause of occurrence of the defective processing state when the defective processing state occurs in the processing device. According to the technique disclosed in Patent Document 1, the judgment result of each inspection apparatus and the image used for the judgment are stored in the database of the management apparatus, and integrated management is performed. The management device has an input unit such as a keyboard and a mouse, and a display unit. Furthermore, when the operator performs an input operation of specifying the analysis target substrate through the input unit, the determination result and the image in each inspection apparatus are simultaneously displayed on the display unit. Thereby, the operator can analyze the cause of the defective processing state in the processing apparatus while checking the judgment result and the image displayed on the display unit.

However, even in the state in which the judgment result of the processing state is outputted in the judgment based on the image, the final quality of the component mounting board may not be high enough. Therefore, it is required to improve the accuracy of the determination based on the correlation image of the processing state of the processing device, and there is still room for improvement in this respect. [Prior Art Literature] [Patent Literature]

[Patent Document 1] Japanese Patent Laid-Open No. 2006-210729

An object of the present invention is to improve the determination accuracy when determining based on an image related to the processing state of a processing device in a component mounting line.

A component mounting system of one aspect of the present invention includes: a component mounting line including a plurality of processing devices and one or a plurality of inspection devices, wherein the plurality of processing devices includes a component mounting device for producing a component mounting substrate on which the components are mounted, the one Or a plurality of inspection devices inspect the processing state of the above-mentioned processing device; and a management device that manages the above-mentioned component mounting line. In the above-mentioned component mounting device and the above-mentioned inspection device, at least any arbitrary plurality of specific devices respectively include: an imaging unit that acquires an image of the processing state of the processing device; and a first determining unit that determines the processing by detecting the feature quantity of the image. Whether the state is good or not, and outputs the first judgment information indicating the judgment result; the second judgment part uses the image classification method based on artificial intelligence to classify the above-mentioned image into the first category and the second category, and the above-mentioned image In the case of the above-mentioned first category, the processing state is judged to be good, and when the image belongs to the second category, the processing state is judged to be bad, and the second judgment information indicating the judgment result is output; and 3. A determination unit that outputs third determination information indicating a final determination result of whether the processing state is good or not, based on the first determination information and the second determination information. The management device includes: an operation unit for inputting type designation command information, and the type designation command information designates which of the first type and the second type the images corresponding to the plurality of specific devices belong to. ; a class information generation unit that generates class information based on the third determination information corresponding to the most downstream device arranged in the most downstream of the component mounting line among the plurality of specific devices, and the class designation command information, and the class information Indicates which of the first category and the second category the images corresponding to the plurality of specific devices belong to; a learning data generation unit corresponding to the second determination unit of each of the plurality of specific devices generating learning data for associating the image with the category information; and a learning section for learning a classification criterion when the second determination section of the plurality of specific devices classifies the image based on the learning data.

The objects, features and advantages of the present invention are further described by the following detailed description and accompanying drawings.

As shown in FIG. 1 , the component mounting system 100 of the present embodiment includes a component mounting line 10 , a management data storage device 13 , an algorithm storage device 14 , and a management device 15 .

The component mounting line 10 is configured by connecting a plurality of processing apparatuses 11 and a plurality of inspection apparatuses 12 so as to be arranged in a straight line. The plurality of processing apparatuses 11 include a solder printing apparatus 11A, a component mounting apparatus 11B, and a reflow apparatus 11C. The inspection device 12 is disposed on the downstream side of each of the plurality of processing devices 11 , and is a device for inspecting the processing state of the corresponding processing device 11 . The plurality of inspection apparatuses 12 include a solder paste inspection apparatus 12A, a mounting inspection apparatus 12B, and a reflow inspection apparatus 12C. In the component mounting line 10 , the solder printing device 11A, the solder paste inspection device 12A, the component mounting device 11B, the mounting inspection device 12B, the reflow device 11C, and the reflow inspection device 12C are connected in order in a straight line.

In the component mounting line 10, at least any plural of the component mounting apparatus 11B, the solder paste inspection apparatus 12A, the mounting inspection apparatus 12B, and the reflow inspection apparatus 12C are set as specific apparatuses for determining whether the processing state is good or not based on images. For example, the component mounting apparatus 11B and the reflow inspection apparatus 12C are set as specific apparatuses. In addition, a plurality of inspection apparatuses 12 including the solder paste inspection apparatus 12A, the mounting inspection apparatus 12B, and the reflow inspection apparatus 12C may be set as specific apparatuses. In this embodiment, the entire apparatus of the component mounting apparatus 11B, the solder paste inspection apparatus 12A, the mounting inspection apparatus 12B, and the reflow inspection apparatus 12C is set as a specific apparatus. In this case, among the component mounting device 11B, the solder paste inspection device 12A, the mounting inspection device 12B, and the reflow inspection device 12C, which are set as specific devices, the reflow inspection device 12C arranged at the most downstream in the component mounting line 10 is set. for the most downstream device.

The solder printing apparatus 11A is an apparatus for printing solder paste on a circuit board to obtain a printed solder paste board PP. The printed solder paste substrate PP is carried into the solder paste inspection apparatus 12A. The solder paste inspection device 12A is disposed on the downstream side of the solder printing device 11A, and inspects the processing state of the solder printing device 11A. The printed solder paste substrate PP after being inspected by the solder paste inspection apparatus 12A is carried into the component mounting apparatus 11B. The component mounting apparatus 11B is an apparatus for producing a component mounting board PPA which mounts electronic components (hereinafter referred to as "components") on the printed solder paste board PP. This component mounting board PPA is carried into the mounting inspection apparatus 12B. The mounting inspection device 12B is disposed on the downstream side of the component mounting device 11B, and inspects the processing state of the component mounting device 11B. The component mounting board PPA that has been inspected by the mounting inspection apparatus 12B is carried into the reflow apparatus 11C. The reflow apparatus 11C is an apparatus for performing a reflow process in which the solder on the component mounting board PPA is melted and then hardened, thereby obtaining a reflow board PPB. This reflow board PPB is carried into the reflow inspection apparatus 12C. The reflow inspection apparatus 12C is arranged on the downstream side of the reflow apparatus 11C, and inspects the processing state of the reflow apparatus 11C.

Next, the component mounting apparatus 11B will be described with reference to FIG. 1 and FIGS. 2 to 5 . In addition, in FIG. 3, the directional relationship is shown using XY orthogonal coordinates which are orthogonal to each other on the horizontal plane.

The component mounting apparatus 11B includes a mounting machine body 2 , a mounting control unit 4 , and a mounting communication unit 40 . The mounter body 2 constitutes a structural portion that performs a component mounting process and the like for mounting components on the printed solder paste substrate PP when the component mounting substrate PPA is produced. The installation communication unit 40 is an interface for data communication with the management data storage device 13 and the algorithm storage device 14 to be described later. The mounting control unit 4 controls the parts mounting process of the mounting machine body 2 and the like, and controls the data communication of the mounting communication unit 40 .

The mounter body 2 includes a body frame 21 , a conveyor 23 , a parts supply unit 24 , a head unit 25 , and a substrate support unit 28 .

The main body frame 21 is a structure in which each part constituting the mounting machine main body 2 is arranged, and is formed in a substantially rectangular shape when viewed from a direction (vertical direction) orthogonal to both the X-axis direction and the Y-axis direction. The conveyor 23 extends in the X-axis direction and is arranged on the main body frame 21 . The conveyor 23 conveys the printed solder paste substrate PP in the X-axis direction. The printed solder paste substrate PP conveyed on the conveyor 23 is positioned by the substrate support unit 28 at a predetermined work position (a component mounting position where components are mounted on the printed solder paste substrate PP). The substrate support unit 28 positions the printed solder paste substrate PP on the conveyor 23 by supporting the printed solder paste substrate PP from the lower side.

The parts supply unit 24 is arranged on the area portion of each of both ends of the body frame 21 in the Y-axis direction across the conveyor 23 . In the body frame 21, an area for mounting the parts supply units 24 in a state where a plurality of parts supply units 24 are arranged side by side is divided, and the installation position of each feeder 24F is divided for each part held by the mounting head 251 included in the head unit 25 to be described later. . The feeder 24F is detachably attached to the parts supply unit 24 . The feeder 24F is a device for performing parts supply processing for supplying parts. The feeder 24F is not particularly limited as long as it can hold a plurality of parts and can supply the held parts to a predetermined parts supply position set in the feeder, and is, for example, a belt feeder. The tape feeder is a feeder configured by including a reel on which a parts storage tape for storing parts at predetermined intervals is wound, and feeding the parts by feeding out the parts storage tape from the reel.

The head unit 25 is held by the moving frame 27 . The main body frame 21 is provided with a fixed rail 261 extending in the Y-axis direction, and a ball screw shaft 262 rotatably driven by the Y-axis servo motor 263 . The moving frame 27 is disposed on the fixed rail 261 , and the nut portion 271 provided on the moving frame 27 is screwed to the ball screw shaft 262 . Moreover, in the moving frame 27, the guide member 272 extended in the X-axis direction, and the ball screw shaft 273 driven by the X-axis servomotor 274 are arrange|positioned. The guide member 272 is held by the head unit 25 so as to be movable, and the nut portion provided in the head unit 25 is screwed to the ball screw shaft 273 . Then, the moving frame 27 is moved in the Y-axis direction by the operation of the Y-axis servo motor 263 , and the head unit 25 is moved in the X-axis direction relative to the moving frame 27 by the operation of the X-axis servo motor 274 . . That is, the head unit 25 can move in the Y-axis direction along with the movement of the moving frame 27 , and can move in the X-axis direction along the moving frame 27 . The head unit 25 is movable between the parts supply unit 24 and the printed solder paste substrate PP supported by the substrate support unit 28 . The head unit 25 executes a component mounting process for mounting components on the printed solder paste substrate PP by moving between the component supply unit 24 and the printed solder paste substrate PP.

As shown in FIG. 4 , the head unit 25 includes a plurality of mounting heads 251 . Each mounting head 251 has a suction nozzle 2511 attached to the distal end (lower end). The suction nozzle 2511 is a nozzle that can suction and hold the parts supplied by the feeder 24F. The suction nozzle 2511 performs parts suction processing for suctioning parts. The adsorption nozzle 2511 can be communicated with any one of the negative pressure generating device, the positive pressure generating device and the atmosphere through an electric switching valve. That is, by supplying a negative pressure to the suction nozzle 2511, a component can be suctioned and held by the suction nozzle 2511, and thereafter, the suction and holding of the component can be released by supplying a positive pressure. Each mounting head 251 performs a component mounting process corresponding to each of a plurality of target mounting positions set for the printed solder paste substrate PP to mount the components adsorbed and held by the suction nozzles 2511 on the printed solder paste substrate PP. The component mounting substrate PPA is obtained by performing the component mounting process on the printed solder paste substrate PP with each mounting head 251 .

Each mounting head 251 is movable up and down relative to the frame of the head unit 25 in the Z-axis direction (vertical direction), and is rotatable around a head axis extending in the Z-axis direction. Each mounting head 251 can be moved up and down in the Z-axis direction between a suctionable position where the parts can be suctioned and held by the suction nozzle 2511 and a retracted position on the upper side with respect to the suctionable position. That is, when the parts are adsorbed and held by the adsorption nozzles 2511, the respective mounting heads 251 are lowered from the retracted positions to the adsorption-capable positions, and the parts are adsorbed and held at the adsorption-capable positions. On the other hand, each of the mounting heads 251 after sucking and holding the components ascends from the suckable position to the retracted position. Further, each mounting head 251 can be raised and lowered in the Z-axis direction between a mountable position and the above-mentioned retracted position, and the above-mentioned mountable position is a predetermined position for mounting the parts adsorbed and held by the suction nozzles 2511 on the printed solder paste substrate PP. The location of the target piggyback location.

As shown in FIGS. 2 and 3 , the mounting machine body 2 further includes a mounting imaging unit 3 . The imaging unit 3 is installed to perform an imaging operation of photographing a photographic subject, and acquire a photographed image. The mounted imaging unit 3 includes a first imaging unit 31 , a second imaging unit 32 , and a third imaging unit 33 .

The first imaging unit 31 is provided on the body frame 21 between the parts supply unit 24 and the conveyer 23 , and includes, for example, a CMOS (Complementary Metal-Oxide-Semiconductor), a CCD (Charged-Coupled Device, a charge coupling element) and other imaging elements of the camera. The first imaging unit 31 photographs the parts adsorbed and held by the adsorption nozzles 2511 of the respective mounting heads 251 from the lower side while the head unit 25 is moving from the parts supply unit 24 toward the printed solder paste substrate PP supported by the substrate support unit 28 . , to obtain the first adsorption image GA1. The first suction image GA1 is an image showing the processing state of the parts suction processing by the suction nozzle 2511 . The first suction image GA1 is an image that can confirm the processing status of the parts suction processing. The processing status of the parts suction processing is, for example, the posture of the parts suctioned by the suction nozzle 2511 and the offset of the suction position of the parts relative to the suction nozzle 2511. Wait. The first suction image GA1 is input to the mounting control unit 4 to be described later, and is used as a reference when determining whether or not the processing state is good by the first suction determination unit 46 and the second suction determination unit 47 .

The second imaging unit 32 is disposed in the head unit 25, and is a camera including, for example, an imaging element such as a CMOS and a CCD. The second imaging unit 32 images the parts held by suction by the suction nozzles 2511 of the respective mounting heads 251 from the side, and acquires a second suction image GA2. The second suction image GA2 is an image showing the processing state of the parts suction processing by the suction nozzle 2511 . The second suction image GA2 is an image for confirming the processing state of the parts suction processing. The processing status of the parts suction processing is, for example, the posture of the parts suctioned by the suction nozzle 2511 and the like. The second suction image GA2 is input to the mounting control unit 4 and is used as a reference when determining whether or not the processing state is good by the first suction determination unit 46 and the second suction determination unit 47 .

The third imaging unit 33 is disposed in the head unit 25, and is a camera including, for example, an imaging element such as a CMOS and a CCD. The third imaging unit 33 captures various marks attached to the upper surface of the printed solder paste substrate PP supported by the substrate support unit 28 from the upper side when each mounting head 251 performs the component mounting process. The amount of positional shift of the printed solder paste substrate PP with respect to the origin coordinates is detected by recognizing the mark on the printed solder paste substrate PP with the third imaging unit 33 .

The installation control unit 4 includes a CPU (Central Processing Unit, central processing unit), a ROM (Read Only Memory) for storing control programs, a RAM (Random Access Memory, random access memory) used as a work area of the CPU, and the like. The installation control unit 4 controls the operation of each component of the installation machine main body 2 by executing the control program stored in the ROM with the CPU, and controls the data communication operation of the installation communication unit 40, thereby executing various calculation processes. As shown in FIG. 2 , the mounting control unit 4 includes a communication control unit 41 , a board conveyance control unit 42 , a parts supply control unit 43 , a head control unit 44 , an imaging control unit 45 , a first adsorption determination unit 46 , and a second adsorption determination unit 47 and the third adsorption determination unit 48 are configured as main functions.

The communication control unit 41 controls data communication between the component mounting device 11B and the management data storage device 13 and data communication between the component mounting device 11B and the algorithm storage device 14 by controlling the mounting communication unit 40 . The installation communication unit 40 transmits the parts adsorption data set DSPP to the management data storage device 13 , and the parts adsorption data set DSPP is a combination of the first adsorption image GA1 and the second adsorption image GA2 and the first adsorption image GA1 and the second adsorption image GA2 to be described later. Output from the third adsorption determination part 48 . 3. Adsorption judgment information DA3 to establish a relationship. In addition, the installation communication unit 40 receives and stores an inspection algorithm to be described later in the algorithm memory device 14 .

The board conveyance control unit 42 controls the conveyance operation of the printed solder paste board PP by the conveyor 23 . The parts supply control unit 43 controls the parts supply process of each of the plurality of feeders 24F arranged in the parts supply unit 24 . The head control unit 44 controls the mounting head 251 by controlling the head unit 25 . Thereby, the head control unit 44 causes the mounting head 251 to execute the component mounting process in correspondence with the plurality of target mounting positions set for the printed solder paste substrate PP, so that the components adsorbed and held by the adsorption nozzles 2511 are mounted on the printed solder paste substrate PP. Solder paste substrate PP. The imaging control unit 45 controls the imaging operation performed by the first imaging unit 31 , the second imaging unit 32 , and the third imaging unit 33 constituting the mounted imaging unit 3 .

The first suction determination unit 46 determines whether or not the processing state of the parts suction processing by the suction nozzle 2511 is good by detecting the feature values of the first suction image GA1 and the second suction image GA2, and outputs the first suction judgment indicating the result of the determination. Information DA1. The above-mentioned feature amount is represented by the size, area, and the like of the region portion of the first suction image GA1 and the second suction image GA2 for confirming the posture of the parts suctioned by the suction nozzle 2511, the offset of the suction position, and the like.

The second suction determination unit 47 uses an image classification method based on AI (Artificial Intelligence) to determine whether or not the processing state of the parts suction processing by the suction nozzle 2511 is good. Specifically, the second suction determination unit 47 classifies the first suction image GA1 and the second suction image GA2 into the first category and the second category, and when it belongs to the first category, determines the suction of the parts by the suction nozzle 2511 The processing status of the processing is good, and when it belongs to the second category, the processing status is judged to be bad. Further, the second adsorption determination unit 47 outputs second adsorption determination information DA2 indicating the above determination result obtained by using an image classification method based on AI (Artificial Intelligence).

The third suction judgment unit 48 outputs third suction judgment information DA3 based on the first suction judgment information DA1 and the second suction judgment information DA2. The third suction judgment information DA3 indicates whether the processing state of the parts suction processing by the suction nozzle 2511 is good or not. judgement result.

As described above, in the parts mounting apparatus 11B, the third suction determination unit 48 performs the parts suction processing of the suction nozzle 2511 based on the determination results of the first suction determination unit 46 and the second suction determination unit 47 whose determination methods are different from each other. The final determination of whether the processing state is good or not. Thereby, the determination accuracy of the final determination of whether or not the processing state is good can be improved.

When the determination result of whether the processing state indicated by the first adsorption determination information DA1 is good or not is consistent with the determination result of whether the processing state indicated by the second adsorption determination information DA2 is good, the third adsorption determination unit 48 outputs an output indicating the matching determination. The result is the third adsorption determination information DA3 of the final determination result. As a result, the determination accuracy of the final determination of whether the processing state of the parts suction processing by the suction nozzle 2511 is good can be maintained at a high level.

For example, when the determination results indicated by the first adsorption determination information DA1 and the second adsorption determination information DA2 indicate that the processing state is good, and the determination results are the same, the third adsorption determination unit 48 outputs the third adsorption determination indicating that the processing state is good Information DA3.

On the other hand, when the determination results indicated by the first adsorption determination information DA1 and the second adsorption determination information DA2 each indicate that the processing state is poor, and the determination results match, the third adsorption determination unit 48 outputs a third adsorption determination unit 48 indicating that the processing state is poor. Adsorption judgment information DA3. In this case, the third suction determination unit 48 may be configured to output, as the third suction determination information DA3, information indicating the determination result of the visual determination performed by the operator using the first suction image GA1 and the second suction image GA2. That is, when both the first adsorption determination information DA1 and the second adsorption determination information DA2 are information indicating the determination result of the poor processing state, the operator may also use the first adsorption image GA1 and the second adsorption image GA2 for visual determination. . And the 3rd adsorption|suction determination part 48 outputs the information which shows the determination result of the visual determination of an operator as 3rd adsorption|suction determination information DA3.

When the judgment result of whether the processing state indicated by the first adsorption judgment information DA1 is good and the judgment result of whether the processing status indicated by the second adsorption judgment information DA2 is not good is inconsistent, the third adsorption judgment unit 48 outputs the final judgment result. This is the third adsorption judgment information DA3 of poor processing state. When the judgment results indicated by the first suction judgment information DA1 and the second suction judgment information DA2 are inconsistent, any one of the first suction judgment information DA1 and the second suction judgment information DA2 indicates that the judgment result is a poor processing state. information. In this case, it is impossible to completely deny the possibility that the processing state of the parts suction processing of the suction nozzle 2511 is poor. Therefore, when the determination results indicated by the first adsorption determination information DA1 and the second adsorption determination information DA2 do not match, the third adsorption determination unit 48 outputs the third adsorption determination information DA3 indicating that the final determination result is poor processing. Thereby, it is possible to more strictly suppress the deterioration of the quality of the parts mounting substrate PPA due to the poor processing state of the parts suction process by the suction nozzle 2511 .

In addition, the third adsorption determination unit may be configured so that when the determination result of whether the processing status indicated by the first adsorption determination information DA1 is good and the determination result of whether the processing status indicated by the second adsorption determination information DA2 are good or not do not coincide with each other, the third adsorption determination section 48 outputs third adsorption determination information DA3 indicating that the final determination result is that the processing state is good. In this case, it is possible to prevent the determination criteria of the final determination by the third adsorption determination unit 48 from being set too strict.

Furthermore, when the determination results indicated by the first adsorption determination information DA1 and the second adsorption determination information DA2 do not match, the third adsorption determination unit 48 may output an output indicating that the operator uses the first adsorption image GA1 and the second adsorption image GA1. 2. The information of the judgment result of the visual judgment of the suction image GA2 is regarded as the third suction judgment information DA3. That is, when the determination results indicated by the first adsorption determination information DA1 and the second adsorption determination information DA2 do not match, the operator may perform visual determination using the first adsorption image GA1 and the second adsorption image GA2. And the 3rd adsorption|suction determination part 48 outputs the information which shows the determination result of the visual determination of an operator as 3rd adsorption|suction determination information DA3.

The third suction determination information DA3 output from the third suction determination unit 48 is associated with the first suction image GA1 and the second suction image GA2, and is set as a part suction data set DSPP. The parts adsorption data set DSPP is transmitted to the management data storage device 13 via the installation communication unit 40 .

Next, the component mounting operation of the component mounting apparatus 11B will be described with reference to the flowchart of FIG. 5 .

When the operator operates to input a command signal to start the component mounting operation of the components on the printed solder paste substrate PP, the component mounting apparatus 11B starts the component mounting operation. First, the printed solder paste substrate PP is conveyed on the conveyor 23 and positioned at a predetermined component mounting position. Then, the parts supply control unit 43 controls the parts supply process of the feeder 24F.

The head control section 44 moves the head unit 25 so that the suction nozzle 2511 is positioned above the parts supplied by the feeder 24F. When the suction nozzle 2511 is positioned above the component, the head control unit 44 lowers the suction nozzle 2511 . The descending suction nozzle 2511 suctions and holds the parts (step a1). When the suction nozzle 2511 suctions and holds the components, the head control unit 44 raises the suction nozzle 2511 . At this time, the second imaging unit 32 images the parts sucked and held by the suction nozzle 2511 from the side, and acquires the second suction image GA2 (step a2). Furthermore, the head control unit 44 moves the head unit 25 so that the suction nozzle 2511 is positioned above the component mounting position of the printed solder paste substrate PP on the conveyor 23 . At this time, the 1st imaging part 31 images the part adsorbed and held by the adsorption nozzle 2511 from below, and acquires 1st adsorption|suction image GA1 (step a2).

When the first imaging unit 31 acquires the first adsorption image GA1 and the second imaging unit 32 acquires the second adsorption image GA2, the first adsorption determination unit 46 detects the first adsorption image GA1 and the second adsorption image GA2 by detecting the first adsorption image GA1 and the second adsorption image GA2. The characteristic value of the suction nozzle 2511 determines whether the processing state of the parts suction processing of the suction nozzle 2511 is good, and outputs the first suction judgment information DA1 indicating the judgment result (step a3). Moreover, the 2nd adsorption|suction determination part 47 classifies the 1st adsorption|suction image GA1 and the 2nd adsorption|suction image GA2 into a 1st category and a 2nd category. Then, the second suction determination unit 47 determines whether the processing state is good or not based on the categories to which the first suction image GA1 and the second suction image GA2 belong, and outputs second suction determination information DA2 indicating the result of the determination (step a3).

The third suction determination unit 48 outputs the third suction determination information DA3 indicating the final determination result of whether the processing state of the parts suction processing by the suction nozzle 2511 is good or not based on the first suction determination information DA1 and the second suction determination information DA2 (step a4) .

When the third adsorption determination information DA3 is output from the third adsorption determination unit 48, the head control unit 44 determines whether or not the determination result of the third adsorption determination information DA3 indicates that the processing state of the parts adsorption process of the adsorption nozzle 2511 is good (step a5). When it is determined that the processing state of the parts suction process is good (YES in step a5), the head control unit 44 lowers the suction nozzle 2511 arranged above the parts mounting position, and applies the printed solder paste substrate PP to the suction nozzle 2511. The component mounting operation of the component is performed (step a6). In this way, components can be mounted on the printed solder paste substrate PP.

On the other hand, when the result of determination is that the processing state of the parts suction processing is not good (NO in step a5), the head control unit 44 controls the operation of the head unit 25 to discard the parts suctioned and held by the suction nozzle 2511. The parts discarding action of (step a7). In this way, by discarding the parts in the poor processing state of the parts adsorption treatment, it is possible to suppress the occurrence of defects when the parts are mounted on the printed solder paste substrate PP.

Next, the inspection apparatus 12 including the solder paste inspection apparatus 12A, the mounting inspection apparatus 12B, and the reflow inspection apparatus 12C will be described with reference to FIGS. 1 , 6 and 7 . The inspection apparatus 12 includes an inspection communication unit 121 , an inspection imaging unit 122 , and an inspection control unit 123 .

The inspection communication unit 121 is an interface for data communication with the management data storage device 13 and the algorithm storage device 14 to be described later.

The inspection imaging unit 122 is, for example, a camera including an imaging element such as a CMOS and a CCD. The inspection imaging unit 122 acquires an inspection image GI of an image showing the processing state of the corresponding processing device 11 . For example, the inspection imaging unit 122 of the solder paste inspection apparatus 12A acquires, as the inspection image GI, an image of the printed solder paste substrate PP that can confirm the processing state of the printing process of the solder printing apparatus 11A. Similarly, the inspection imaging unit 122 of the mounting inspection apparatus 12B acquires an image of the component mounting substrate PPA that can confirm the processing state of the component mounting process of the component mounting apparatus 11B as the inspection image GI. Moreover, the inspection imaging part 122 of the reflow inspection apparatus 12C acquires the image of the reflow board PPB which can confirm the processing state of the reflow process of the reflow apparatus 11C, as an inspection image GI.

The inspection control unit 123 includes a CPU, a ROM for storing a control program, a RAM used as a work area of the CPU, and the like. The inspection control unit 123 controls the inspection communication unit 121 and the inspection imaging unit 122 by executing the control program stored in the ROM by the CPU, and executes various arithmetic processing. The inspection control unit 123 includes a communication control unit 1231 , an imaging control unit 1232 , a first inspection determination unit 1233 , a second inspection determination unit 1234 , and a third inspection determination unit 1235 as main functional configurations.

The communication control unit 1231 controls the data communication between the inspection device 12 and the management data storage device 13 and the data communication between the inspection device 12 and the algorithm storage device 14 by controlling the inspection communication unit 121 . The inspection communication unit 121 transmits an inspection data set DS obtained by associating the inspection image GI with the third inspection determination information DI3 output from the third inspection determination unit 1235 described later to the management data storage device 13 . For example, the inspection communication unit 121 of the solder paste inspection device 12A transmits the solder paste inspection data set DSPT obtained by associating the inspection image GI with the third inspection determination information DI3 to the management data storage device 13 . Similarly, the inspection communication unit 121 of the installation inspection device 12B transmits the installation inspection data set DSMT obtained by associating the inspection image GI and the third inspection determination information DI3 to the management data storage device 13 . Further, the inspection communication unit 121 of the reflow inspection apparatus 12C transmits the reflow inspection data set DSRF obtained by associating the inspection image GI with the third inspection determination information DI3 to the management data storage device 13 . In addition, the inspection communication unit 121 receives and stores an inspection algorithm to be described later in the algorithm storage device 14 .

The imaging control unit 1232 controls the imaging operation of the inspection imaging unit 122 .

The first inspection determination unit 1233 determines whether or not the processing state of the corresponding processing device 11 is good by detecting the feature quantity of the inspection image GI, and outputs the first inspection determination information DI1 indicating the determination result. The above-mentioned feature amount is represented by the size, area, etc. of the region portion of the inspection image GI for confirming the processing state of the corresponding processing device 11 .

The second inspection and determination unit 1234 determines whether or not the processing state of the corresponding processing device 11 is good by using an image classification method based on AI (Artificial Intelligence). Specifically, the second inspection determination unit 1234 classifies the inspection image GI into the first category and the second category, and when the inspection image GI belongs to the first category, determines that the processing state of the corresponding processing device 11 is good, and When the inspection image GI belongs to the second category, it is determined that the processing state is not good. Then, the second inspection determination unit 1234 outputs second inspection determination information DI2 indicating the above determination result obtained by using the image classification method based on AI (Artificial Intelligence).

Based on the first inspection determination information DI1 and the second inspection determination information DI2, the third inspection determination unit 1235 outputs the third inspection determination information DI3 indicating the final determination result of whether the processing state of the corresponding processing device 11 is good.

As described above, in the inspection device 12, the third inspection determination unit 1235 determines whether or not the processing state of the corresponding processing device 11 is based on the determination results of the first inspection determination unit 1233 and the second inspection determination unit 1234 whose determination methods are different from each other. Good final judgment. Thereby, the determination accuracy of the final determination of whether or not the processing state is good can be improved.

In the case where the determination result of whether the processing state indicated by the first inspection determination information DI1 is good and the determination result of whether the processing state indicated by the second inspection determination information DI2 is good is consistent, the third inspection determination unit 1235 outputs a result indicating the agreement. The judgment result is the third inspection judgment information DI3 of the final judgment result. Thereby, the determination accuracy of the final determination of whether the processing state of the corresponding processing device 11 is good can be maintained at a high level.

For example, when the determination results indicated by the first inspection determination information DI1 and the second inspection determination information DI2 indicate that the processing status is good, and the processing results are the same, the third inspection determination unit 1235 outputs a third inspection determination indicating that the processing status is good Information DI3.

On the other hand, when the determination results indicated by the first inspection determination information DI1 and the second inspection determination information DI2 indicate that the processing state is not good, and the processing results are the same, the third inspection and determination unit 1235 outputs a third inspection and determination unit 1235 indicating that the processing state is not good. Check judgment information DI3. In this case, the third inspection determination unit 1235 may be configured to output, as the third inspection determination information DI3, information indicating the determination result of the visual determination performed by the operator using the inspection image GI. That is, in the case where both the first inspection determination information DI1 and the second inspection determination information DI2 are information indicating the determination result of the poor processing state, the operator can also perform visual determination using the inspection image GI. Then, the third inspection determination unit 1235 outputs information indicating the determination result of the operator's visual determination as third inspection determination information DI3.

In the case where the judgment result of whether the processing state indicated by the first inspection judgment information DI1 is good and the judgment result of whether the processing status indicated by the second inspection judgment information DI2 is good or not are inconsistent, the third inspection judgment unit 1235 outputs the final judgment result. It is the third inspection judgment information DI3 of the bad processing state. When the judgment results shown in the first inspection judgment information DI1 and the second inspection judgment information DI2 do not match, any one of the first inspection judgment information DI1 and the second inspection judgment information DI2 indicates that the judgment result is a poor processing state. information. In this case, it is impossible to completely deny the possibility that the processing state of the corresponding processing device 11 is bad. Therefore, when the determination results indicated by the first inspection determination information DI1 and the second inspection determination information DI2 do not match, the third inspection determination unit 1235 outputs the third inspection determination information DAI3 indicating that the final determination result is poor processing. Thereby, the quality degradation of the component mounting substrate PPA due to the poor processing state of the corresponding processing apparatus 11 can be more strictly suppressed.

In addition, the third inspection and determination unit may be configured such that when the result of the determination of whether the processing state indicated by the first inspection and determination information DI1 is good and the result of the determination of whether the processing state of the second inspection and determination information DI2 is good or not are inconsistent with each other, the third inspection and determination unit may 1235 outputs third inspection determination information DI3 indicating that the final determination result is that the processing state is good. In this case, it is possible to prevent the determination criteria of the final determination by the third inspection determination unit 1235 from being set too strict.

Furthermore, when the determination results indicated by the first inspection determination information DI1 and the second inspection determination information DI2 do not match, the third inspection determination unit 1235 may output a determination indicating that the operator performs visual determination using the inspection image GI. The information of the result is regarded as the third inspection determination information DI3. That is, in the case where the determination results indicated by the first inspection determination information DI1 and the second inspection determination information DI2 do not match, the operator may perform visual determination using the inspection image GI. Then, the third inspection determination unit 1235 outputs information indicating the determination result of the operator's visual determination as third inspection determination information DI3.

The third inspection determination information DI3 output from the third inspection determination unit 1235 is associated with the inspection image GI, and is set as an inspection data set DS. The examination data set DS is transmitted to the management data storage device 13 via the examination communication unit 121 .

Next, the inspection determination operation of the inspection device 12 will be described with reference to the flowchart of FIG. 7 .

The inspection apparatus 12 starts the inspection and determination operation when the substrate carried out from the corresponding processing apparatus 11 is carried in (step b1 ). First, the inspection imaging unit 122 acquires the inspection image GI (step b2).

When the inspection imaging unit 122 acquires the inspection image GI, the first inspection determination unit 1233 determines whether the processing state of the corresponding processing device 11 is good or not by detecting the feature quantity of the inspection image GI, and outputs the first inspection indicating the determination result. Judgment information DI1 (step b3). Further, the second inspection determination unit 1234 classifies the inspection image GI into the first category and the second category. Then, the second inspection determination unit 1234 determines whether the processing state is good or not based on the category to which the inspection image GI belongs, and outputs the second inspection determination information DI2 indicating the determination result (step b3).

Based on the first inspection determination information DI1 and the second inspection determination information DI2, the third inspection determination unit 1235 outputs the third inspection determination information DI3 indicating the final determination result of whether the processing state of the corresponding processing device 11 is good or not (step b4).

When the third inspection determination information DI3 is output from the third inspection determination unit 1235, the inspection control unit 123 determines whether the determination result of the third inspection determination information DI3 indicates that the processing state of the corresponding processing device 11 is good (step b5). When the result of determination is that the processing state is good (YES in step b5 ), the inspection control unit 123 unloads the inspected substrate from the inspection apparatus 12 to the downstream side (step b6 ).

On the other hand, when the result of determination is that the processing state is not good (NO in step b5 ), the inspection control unit 123 executes a substrate discarding operation for discarding the substrate after inspection (step b7 ). At this time, for example, when the third inspection determination information DI3 indicating that the determination result is poor processing state is outputted from the mounted inspection device 12B, if the target part in the poor processing state can be replaced with a new one, the operator may discard it. The defective target parts are replaced with new parts to repair the parts mounting board PPA. On the other hand, in the case where the defective object parts cannot be replaced with new parts, the operator discards the parts mounting board PPA.

Next, the management data storage device 13 will be described with reference to FIG. 1 and FIG. 8 . The management data storage device 13 and the component mounting device 11B are connected so as to be capable of data communication. In addition, the management data storage device 13 is connected to each of the solder paste inspection device 12A, the mounting inspection device 12B, and the reflow inspection device 12C constituting the plurality of inspection devices 12 so as to be capable of data communication. As described above, the component suction data set DSPP output from the component mounting device 11B is input to the management data storage device 13 . Also, the solder paste inspection data set DSPT output from the solder paste inspection device 12A, the mounting inspection data set DSMT output from the mounting inspection device 12B, and the reflow inspection data set DSRF output from the reflow inspection device 12C are also input to the management Data storage device 13 .

The management data storage device 13 is a device for storing and storing the management data DM shown in FIG. 8 . The management data DM is data associated with the substrate information DPP, the parts information DP, the solder paste inspection data set DSPT, the part adsorption data set DSPP, the mounting inspection data set DSMT, and the reflow inspection data set DSRF. The data group of the management data DM is stored in the management data storage device 13 .

The substrate information DPP is used to specify the information of the printed solder paste substrate PP, the component mounting substrate PPA and the reflow substrate PPB. In the substrate information DPP, information such as a substrate ID (Identification) is registered as information for specifying each substrate. The part information DP is information for specifying the parts used when the part mounting board PPA is produced in the part mounting apparatus 11B. In the part information DP, information such as a part name inherent to a part, a part type indicating the type of the part, the external dimension of the part, and the like are registered as information for specifying the part.

As described above, the solder paste inspection data set DSPT is an inspection data set output from the solder paste inspection apparatus 12A, and is a data set obtained by associating the inspection image GI with the third inspection determination information DI3. As described above, the parts suction data set DSPP is a data set output from the component mounting apparatus 11B, and is a data set that associates the first suction image GA1 and the second suction image GA2 with the third suction determination information DA3. As described above, the on-board inspection data set DSMT is an inspection data set output from the on-board inspection apparatus 12B, and is a data set in which the inspection image GI and the third inspection determination information DI3 are associated. As described above, the reflow inspection data set DSRF is an inspection data set output from the reflow inspection apparatus 12C, and is a data set obtained by associating the inspection image GI with the third inspection determination information DI3.

Next, the management device 15 will be described with reference to FIG. 1 and FIGS. 9 and 10 . The management device 15 is a device for managing the component mounting line 10 . The management device 15 is connected to the management data storage device 13 and the algorithm storage device 14 so as to be capable of data communication. The management device 15 can read into the data group of the management data DM stored and memorized in the management data storage device 13 . In addition, the management device 15 may input the inspection algorithm IAL generated by the learning unit 1545 described later to the algorithm memory device 14 .

The algorithm memory device 14 is a device for storing the inspection algorithm IAL from the management device 15 . The algorithm memory device 14 and the component mounting device 11B are connected so as to be capable of data communication. In addition, the algorithm memory device 14 is connected to each of the solder paste inspection device 12A, the mounting inspection device 12B, and the reflow inspection device 12C constituting the plurality of inspection devices 12 so as to be capable of data communication. Therefore, the component mounting device 11B can read the inspection algorithm IAL stored in the algorithm memory device 14 . In the component mounting apparatus 11B, the second suction determination unit 47 classifies the first suction image GA1 and the second suction image GA2 into the first category and the second category based on the inspection algorithm IAL, thereby determining whether the processing state of the suction nozzle 2511 is not. good. Likewise, each inspection device 12 can read the inspection algorithm IAL stored in the algorithm memory device 14 . In each inspection device 12, the second inspection determination unit 1234 classifies the inspection image GI into the first category and the second category based on the inspection algorithm IAL, and determines whether or not the processing state of the corresponding processing device 11 is good.

The management device 15 includes a management communication unit 151 , a display unit 152 , an operation unit 153 , and a management control unit 154 .

The management communication unit 151 is an interface for data communication with the management data storage device 13 and the algorithm storage device 14 . The management communication unit 151 acquires the data group of the management data DM from the management data storage device 13 , and transmits the check algorithm IAL to the algorithm storage device 14 .

The display unit 152 includes, for example, a liquid crystal display or the like. The display unit 152 displays various images such as the first suction image GA1 , the second suction image GA2 , and the inspection image GI included in the data group of the management data DM according to a request from the category information generation unit 1543 described later.

The operation unit 153 includes a keyboard, a mouse, a touch panel provided on the display unit 152 , and the like. The operation unit 153 accepts input operations of various commands by the operator. The operation unit 153 is for the operator to input the category designation command information DCC. The category designation command information DCC is information for specifying which of the first category and the second category the first adsorption image GA1 and the second adsorption image GA2 belong to, the first adsorption image GA1 and the second adsorption image GA2 is an image corresponding to the component mounting device 11B, and constitutes an image of the component adsorption data set DSPP included in the management data DM. The category designation command information DCC is information for specifying which of the first category and the second category the inspection image GI belongs to. The inspection image GI is an image corresponding to each inspection device 12 and constitutes management data. The images of the solder paste inspection data set DSPT, the mounting inspection data set DSMT and the reflow inspection data set DSRF included in the DM.

The operation part 153 corresponds to the display of the first suction image GA1 , the second suction image GA2 , and the inspection image GI in the display part 152 , and accepts the input of the type designation command information DCC. Thereby, the operator can visually determine which of the first category and the second category the respective images belong to while checking each image displayed on the display unit 152 , and input from the operation unit 153 corresponding to the visual determination. The class specifies the command information DCC.

The management control unit 154 includes a CPU, a ROM for storing a control program, a RAM used as a work area of the CPU, and the like. The management control unit 154 controls the management communication unit 151 and the display unit 152 by executing the control program stored in the ROM by the CPU, and executes various calculation processes. The management control unit 154 includes a communication control unit 1541, a display control unit 1542, a category information generation unit 1543, a learning data generation unit 1544, and a learning unit 1545 as main functional configurations.

The communication control part 1541 controls the data communication between the management device 15 and the management data storage device 13 and the data communication between the management device 15 and the algorithm storage device 14 by controlling the management communication part 151 . The display control unit 1542 controls the display unit 152 .

The category information generating unit 1543 generates category information DCT indicating that the first suction image GA1 and the second suction image GA2 corresponding to the component mounting apparatus 11B, and the inspection image GI corresponding to each inspection apparatus 12 belong to the first category with which of the 2 categories. The category information generation unit 1543 generates category information DCT in association with the first suction image GA1 and the second suction image GA2 included in the data group of the management data DM, and the entire image of the inspection image GI. At this time, the classification information generation unit 1543 generates the classification information DCT based on the third inspection determination information DI3 and the classification designation command information DCC inputted through the operation unit 153 , and the third inspection determination information DI3 is configured to be arranged most downstream in the component mounting line 10 . The reflow inspection data set DSRF corresponding to the reflow inspection device 12C (specific inspection device).

The learning data generation unit 1544 generates a result obtained by associating the first suction image GA1 and the second suction image GA2 included in the data group of the management data DM with the category information DCT in correspondence with the second suction determination unit 47 of the component mounting apparatus 11B The learning materials DLN. In addition, the learning data generating unit 1544 generates learning data DLN obtained by associating the inspection image GI included in the data group of the management data DM with the category information DCT in correspondence with the second inspection determination unit 1234 of each inspection apparatus 12 .

The learning unit 1545 performs machine learning on the classification criteria used by the second suction determination unit 47 to classify the first suction image GA1 and the second suction image GA2 based on the learning data DLN corresponding to the component mounting device 11B, and generates an inspection algorithm IAL, The inspection algorithm IAL represents an image classification procedure according to the classification criterion. In addition, the learning unit 1545 performs machine learning on the classification criteria when the second inspection determination unit 1234 classifies the inspection image GI based on the learning data DLN corresponding to each inspection apparatus 12, and generates an inspection algorithm IAL, which indicates a basis The image classification program of the classification benchmark. The inspection algorithm IAL generated by the learning unit 1545 is stored in the algorithm memory device 14 .

The method by which the learning unit 1545 performs the machine learning is not particularly limited, and for example, a method using a neural network (Neural Network) can be exemplified. The neural network has a structure that simulates the structure of the human brain, and is composed of multiple layers of logic circuits that simulate the functions of neurons (nerve cells) in the human brain. A neural network includes an input layer, a hidden layer, and an output layer as a logic circuit layer.

As described above, the learning data DLN used by the learning unit 1545 to learn the classification criteria when the second suction determination unit 47 of the component mounting device 11B and the second inspection determination unit 1234 of each inspection device 12 perform image classification is learned by learning The data generation unit 1544 generates. The learning data DLN corresponding to the second suction determination unit 47 of the component mounting apparatus 11B is data obtained by associating the first suction image GA1 and the second suction image GA2 included in the data group of the management data DM with the category information DCT . The learning data DLN corresponding to the second inspection determination unit 1234 of each inspection apparatus 12 is data obtained by associating the inspection image GI included in the data group of the management data DM with the category information DCT.

The category information DCT constituting the learning data DLN is usually set by the operator inputting the category designation command information DCC through the operation unit 153 . That is, in general, the operator needs to perform the classification designation command information DCC corresponding to the entire image of the first suction image GA1, the second suction image GA2, and the inspection image GI included in the data group of the management data DM. Enter an action. In this case, in order to improve the determination accuracy of the second suction determination unit 47 of the component mounting device 11B and the second inspection determination unit 1234 of each inspection device 12, the operator must issue a category designation command corresponding to a large number of images. Input operation of information DCC. In this case, the work load is large for the operator.

Therefore, in the present embodiment, in the case where the category information DCT corresponding to the first suction image GA1 and the second suction image GA2 included in the data group of the management data DM and the inspection image GI corresponding to each image is generated At this time, the type information generation unit 1543 refers not only to the type designation command information DCC, but also to the third inspection determination information DI3 corresponding to the reflow inspection apparatus 12C. In this case, the operator does not need to perform the input operation of the category designation command information DCC corresponding to all the images included in the data group of the management data DM. Thereby, the work load of the operator can be reduced, and the category information DCT corresponding to each image included in the data group of the management data DM can be generated. Therefore, the determination accuracy of the second suction determination unit 47 of the component mounting device 11B and the second inspection determination unit 1234 of each inspection device 12 can be improved while reducing the operator's work load.

Assume the following situation: the third inspection determination information DI3 corresponding to the most downstream reflow inspection apparatus 12C is information indicating a good determination result, and the solder paste inspection apparatus 12A and the mounting inspection apparatus 12B other than the reflow inspection apparatus 12C correspond to The third inspection judgment information DI3 and the third suction judgment information DA3 corresponding to the component mounting device 11B are information indicating a good judgment result. In this case, the category information generation unit 1543 generates category information DCT indicating that each of the first and second adsorption images GA1 and GA2 included in the management data DM and the inspection image GI belong to the first category. In this case, the class information generation unit 1543 automatically generates class information DCT indicating that each image belongs to the first class without referring to the class designation command information DCC. Thereby, the work load of the operator can be more reliably reduced, and the category information DCT corresponding to each image included in the data group of the management data DM can be generated.

As described above, for example, when the third inspection determination information DI3 indicating that the determination result is poor processing status is output from the mounted inspection device 12B, if the target parts with poor processing status can be replaced with new ones, the operator may discard them. The defective target parts are replaced with new parts to repair the parts-mounted board PPA operation. In this case, although the third inspection determination information DI3 corresponding to the mounting inspection device 12B is information indicating that the determination result is not good, the repaired parts mounting substrate PPA is sometimes sent to the downstream side of the mounting inspection device 12B, and the most The third inspection judgment information DI3 corresponding to the downstream reflow inspection device 12C indicates a good judgment result. Assume this situation: the third inspection determination information DI3 corresponding to the most downstream reflow inspection device 12C is information indicating a good determination result, and the solder paste inspection device 12A and the mounting inspection device 12B other than the reflow inspection device 12C are At least any one of the corresponding third inspection judgment information DI3 and the third suction judgment information DA3 corresponding to the component mounting device 11B is information indicating a defective judgment result.

In this case, the category information generating unit 1543 generates the images indicating that the first suction image GA1 and the second suction image GA2 and the inspection image GI included in the management data DM are designated by the type designation command information DCC. The class information DCT of the class. Thereby, when the operator has performed the work of repairing the parts-mounted board PPA, etc., the class corresponding to each image included in the data group of the management data DM can be generated based on the class designation command information DCC designated by the operator. Info DCT.

In addition, when the third inspection determination information DI3 corresponding to the most downstream reflow inspection device 12C is information indicating a defective determination result, the solder printing device 11A, the component mounting device 11B and the reflow device on the upstream side of the reflow inspection device 12C The processing state of at least one of the processing devices 11 in the welding device 11C may be poor. Assume this situation: the third inspection determination information DI3 corresponding to the most downstream reflow inspection device 12C is information indicating the determination result of the defectiveness.

In this case, the category information generating unit 1543 generates the images indicating that the first suction image GA1 and the second suction image GA2 and the inspection image GI included in the management data DM are designated by the type designation command information DCC. The class information DCT of the class. Thereby, when the processing state of each processing device 11 may be bad, the category information DCT corresponding to each image included in the data group of the management data DM can be generated based on the category specification command information DCC specified by the operator . In this case, the operator can only focus on checking the image corresponding to the situation that the processing state of each processing device 11 may be bad, visually determine which of the first category and the second category the image belongs to, and automatically. The operation unit 153 inputs the type designation command information DCC corresponding to the visual judgment.

Next, the learning process of the management apparatus 15 is demonstrated with reference to the flowchart of FIG.

The management device 15 starts the learning process by acquiring the data group of the management data DM stored and memorized by the management data storage device 13 through the management communication unit 151 (step c1). First, the category information generation unit 1543 determines the third inspection corresponding to the most downstream reflow inspection device 12C for each management data DM uniquely identified by the board information DPP and the component information DP included in the data group of the management data DM It is determined whether or not the information DI3 is information indicating a good determination result (step c2).

When the third inspection determination information DI3 corresponding to the reflow inspection device 12C is information indicating a good determination result (Yes in step c2), the category information generation unit 1543 determines the reflow inspection for each management data DM. Whether the third inspection judgment information DI3 corresponding to the solder paste inspection device 12A and the mounting inspection device 12B other than the device 12C, and the third suction judgment information DA3 corresponding to the component mounting device 11B are all information indicating a good judgment result (step c3).

When the third inspection determination information DI3 corresponding to the solder paste inspection device 12A and the mounting inspection device 12B, and the third suction determination information DA3 corresponding to the component mounting device 11B are all information indicating a good determination result (in step c3: Yes), the category information generating unit 1543 generates category information DCT indicating that each of the first and second adsorption images GA1 and GA2 included in the management data DM and the inspection image GI belong to the first category (step c4 ). ).

When the third inspection determination information DI3 corresponding to the reflow inspection device 12C is information indicating a defective determination result (NO in step c2), the process proceeds to step c5. In addition, at least any one of the third inspection determination information DI3 corresponding to the solder paste inspection device 12A and the mounting inspection device 12B and the third suction determination information DA3 corresponding to the component mounting device 11B is information indicating a defective determination result If (NO in step c3 ), the process also proceeds to step c5 .

In step c5, the category information generating unit 1543 outputs a display request signal to the display control unit 1542, the display request signal requesting the display of the first suction image GA1 and the second suction image GA2 associated with the third inspection determination information DI3, and each of them. In the inspection image GI, the third inspection determination information DI3 corresponds to the reflow inspection apparatus 12C, and indicates that the determination result is defective. The display control part 1542 which received the said display request signal makes the display part 152 display the 1st adsorption|suction image GA1, the 2nd adsorption|suction image GA2, and each image of each inspection image GI. Then, the operation unit 153 accepts the input of the category designation command information DCC by the operator in correspondence with the display of the first suction image GA1 and the second suction image GA2 and each inspection image GI in the display unit 152 . While checking each image displayed on the display unit 152 , the operator visually determines which of the first category and the second category the image belongs to, and inputs a category designation command corresponding to the visual determination from the operation unit 153 . Info DCC.

After the above-mentioned display request signal is output, the category information generation unit 1543 determines whether or not category designation command information DCC has been input via the operation unit 153 (step c6). When the category designation command information DCC is input (Yes in step c6 ), the category information generation unit 1543 generates the first suction image GA1 and the second suction image GA2 and the inspection image which are included in the management data DM. Each image of the GI belongs to the class information DCT of the class designated by the class designation command information DCC (step c7).

When the category information DCT is generated by the category information generation unit 1543 in steps c4 and c7, the learning data generation unit 1544 generates the data group included in the management data DM corresponding to the second suction determination unit 47 of the component mounting apparatus 11B. The learning data DLN obtained by associating the first adsorption image GA1 and the second adsorption image GA2 with the category information DCT (step c8 ). Further, the learning data generating unit 1544 generates learning data DLN in which the inspection image GI included in the data group of the management data DM and the category information DCT are associated in correspondence with the second inspection determining unit 1234 of each inspection apparatus 12 (step c8). .

When the learning data DLN is generated by the learning data generating unit 1544, the learning unit 1545 classifies the first suction image GA1 and the second suction image to the second suction determination unit 47 based on the learning data DLN corresponding to the component mounting device 11B Machine learning is performed on the classification benchmark at GA2 (step c9). In addition, the learning unit 1545 performs machine learning on the classification criteria when the second inspection determination unit 1234 classifies the inspection image GI based on the learning data DLN corresponding to each inspection apparatus 12 (step c9). Then, the learning unit 1545 generates an inspection algorithm IAL corresponding to the second suction determination unit 47 of the component mounting device 11B and the second inspection determination unit 1234 of each inspection device 12, and the inspection algorithm IAL indicates that the classification criterion is based on the learned classification. the image classification procedure (step c10).

The inspection algorithm IAL generated by the learning unit 1545 is stored in the algorithm memory device 14 .

In the component mounting device 11B, the second suction determination unit 47 can read the inspection algorithm IAL stored in the algorithm memory device 14, and can classify the first suction image GA1 and the second suction image GA2 based on the inspection algorithm IAL. It is classified into the 1st type and the 2nd type, and it is judged whether the processing state of the suction nozzle 2511 is good. In the parts mounting apparatus 11B, the third suction determination unit 48 determines whether the processing state of the parts suction processing by the suction nozzle 2511 is good or not based on the determination results of the first suction determination unit 46 and the second suction determination unit 47 whose determination methods are different from each other. final judgment. Thereby, the determination accuracy of the final determination of whether or not the processing state is good can be improved.

Further, in each inspection apparatus 12, the second inspection determination unit 1234 can classify the inspection image GI into the first category and the second category based on the inspection algorithm IAL by reading the inspection algorithm IAL stored in the algorithm memory device 14. 2 categories, so as to determine whether the processing state of the corresponding processing device 11 is good. In each inspection device 12, the third inspection determination unit 1235 performs a final determination as to whether the processing state of the corresponding processing device 11 is good or not based on the determination results of the first inspection determination unit 1233 and the second inspection determination unit 1234 whose determination methods are different from each other. determination. Thereby, the determination accuracy of the final determination of whether or not the processing state is good can be improved.

In addition, the above-mentioned specific embodiment mainly includes invention which has the following structures.

A component mounting system of one aspect of the present invention includes: a component mounting line including a plurality of processing devices and one or a plurality of inspection devices, wherein the plurality of processing devices includes a component mounting device for producing a component mounting substrate on which the components are mounted, the one Or a plurality of inspection devices inspect the processing state of the above-mentioned processing device; and a management device that manages the above-mentioned component mounting line. In the above-mentioned component mounting device and the above-mentioned inspection device, at least any arbitrary plurality of specific devices respectively include: an imaging unit that acquires an image of the processing state of the processing device; and a first determining unit that determines the processing by detecting the feature quantity of the image. Whether the state is good or not, and outputs the first judgment information indicating the judgment result; the second judgment part uses the image classification method based on artificial intelligence to classify the above-mentioned image into the first category and the second category, and the above-mentioned image In the case of the above-mentioned first category, the processing state is judged to be good, and when the image belongs to the second category, the processing state is judged to be bad, and the second judgment information indicating the judgment result is output; and 3. A determination unit that outputs third determination information showing a final determination result of whether the processing state is good or not, based on the first determination information and the second determination information. The management device includes: an operation unit for inputting type designation command information, and the type designation command information designates which of the first type and the second type the images corresponding to the plurality of specific devices belong to. ; a class information generation unit that generates class information based on the third determination information corresponding to the most downstream device arranged in the most downstream of the component mounting line among the plurality of specific devices, and the class designation command information, and the class information Displaying which of the first category and the second category the images corresponding to the plurality of specific devices belong to; a learning data generation unit corresponding to the second determination unit of each of the plurality of specific devices generating learning data for associating the image with the category information; and a learning section for learning a classification criterion when the second determination section of the plurality of specific devices classifies the image based on the learning data.

According to this component mounting system, in each specific device on the component mounting line, the third determination unit performs a final determination as to whether the processing state of the processing device is good or not based on the determination results of the first determination unit and the second determination unit whose determination methods are different from each other. determination. Thereby, the determination accuracy of the final determination of whether or not the processing state is good can be improved.

The learning data used by the learning unit to learn the classification criteria when the second determination unit of each specific device performs image classification is generated by the learning data generating unit. The learning data is data obtained by associating images in each specific device with category information. The category information constituting the learning material is usually set by the operator inputting category designation instruction information through the operation unit. That is, generally, the operator needs to perform the input operation of the category designation instruction information corresponding to all the images in each specific device. In this case, in order to improve the determination accuracy of the second determination unit of each specific device, the operator must perform the input operation of the category designation command information corresponding to a large number of images. In this case, the work load is large for the operator.

Therefore, when generating the category information corresponding to each image, the category information generating unit refers not only to the category designation command information, but also to the third determination information corresponding to the most downstream device arranged most downstream in the component mounting line. In this case, the operator does not need to input the category designation instruction information corresponding to all the images in each specific device. Thereby, the workload of the operator can be reduced, and category information corresponding to each image can be generated. Therefore, the determination accuracy of the second determination unit of each specific device can be improved while reducing the operator's workload.

In the above-mentioned component mounting system, the above-mentioned type information generation unit may be configured to generate the above-mentioned plural specific apparatuses when the above-mentioned third determination information corresponding to the above-mentioned most downstream apparatus is information indicating a determination result indicating that the processing state is good The information that the corresponding images belong to the first category is regarded as the category information.

In this aspect, when the third determination information corresponding to the most downstream device arranged at the most downstream of the component mounting line is information indicating a good determination result, the category information generating unit generates category information indicating that each image belongs to the first category . In this case, the category information generation unit automatically generates category information indicating that each image belongs to the first category without referring to the category designation instruction information. Thereby, the work load of the operator can be more reliably reduced, and category information corresponding to each image can be generated.

The component mounting system may be configured as follows: the third determination information corresponding to the most downstream device by the category information generation unit is information indicating the determination result that the processing state is good, and the plurality of specific devices may include the above-mentioned third determination information. When the third determination information corresponding to the other specific devices other than the most downstream device is the information indicating the determination result of the processing state being poor, the generated image indicates that the images corresponding to the plurality of specific devices belong to the category designation command information. The information of the specified category is regarded as the above category information.

For example, when the third judgment information indicating that the judgment result is poor processing state is output from the inspection device arranged on the downstream side of the component mounting device, if the target part of the poor processing state can be replaced with a new one, the operator has At this time, the defective parts can be discarded and replaced with new parts to repair the parts mounting board. In this case, although the third judgment information corresponding to the inspection device on the downstream side of the component mounting device is information indicating the judgment result of defectiveness, the repaired component mounting board may be sent to the downstream side, and the most downstream device may be sent to the downstream side. The corresponding third judgment information shows a good judgment result. Assume this situation: the third judgment information corresponding to the most downstream device is information indicating a good judgment result, and the third judgment information corresponding to the other specific devices other than the most downstream device is information indicating a bad judgment result.

In this case, the category information generating section generates category information indicating that each image belongs to the category specified by the category specifying instruction information. Thereby, when the operator performs the operation of repairing the component mounting board, etc., the class information corresponding to each image can be generated based on the class designation command information designated by the operator.

In the above-mentioned component mounting system, the above-mentioned type information generation unit may be configured to generate the above-mentioned plural specific apparatuses when the above-mentioned third determination information corresponding to the above-mentioned most downstream apparatus is information indicating the determination result of the above-mentioned poor processing state. The information corresponding to the above-mentioned images belonging to the type specified by the above-mentioned type-designating instruction information is regarded as the above-mentioned type information.

When the third judgment information corresponding to the most downstream device is information indicating the judgment result of bad, the processing state of at least one of the processing devices on the upstream side of the most downstream device may be poor. Assume this situation: the third judgment information corresponding to the most downstream device is information indicating a bad judgment result.

In this case, the category information generating section generates category information indicating that each image belongs to the category specified by the category specifying instruction information. Thereby, when the processing state of each processing device may be bad, the type information corresponding to each image can be generated based on the type designation command information designated by the operator. In this case, the operator can only focus on confirming the image corresponding to the situation that the processing state of each processing device may be bad, visually determine which of the first category and the second category the image belongs to, and operate it by himself. The section inputs the category designation instruction information corresponding to the visual judgment.

The component mounting system may be configured as follows: the third determination unit determines whether the processing status indicated by the first determination information is good or not, and the processing status indicated by the second determination information is good or not. When the results match, the above-mentioned third judgment information indicating the matching judgment result as the final judgment result is output.

In this aspect, when the determination result of whether the processing state indicated by the first determination information is good or not is consistent with the determination result of whether the processing state indicated by the second determination information is good or not, the third determination unit outputs a determination indicating the coincidence The result is used as the third judgment information of the final judgment result. Thereby, the determination accuracy of the final determination of whether the processing state of each processing device is good can be maintained at a high level.

The component mounting system may be configured as follows: the third determination unit determines whether the processing status indicated by the first determination information is good or not, and the processing status indicated by the second determination information is good or not. When the results do not match, the third judgment information indicating that the final judgment result is that the processing state is not good is output.

In the case where the determination results indicated by the first determination information and the second determination information do not match, any one of the first determination information and the second determination information is information indicating that the determination result is a poor processing state. In this case, the possibility that the processing state of the processing device is bad cannot be completely denied. Therefore, when the determination results indicated by the first determination information and the second determination information do not match, the third determination unit outputs the third determination information indicating that the final determination result is a poor processing state. Thereby, it is possible to more strictly suppress the deterioration of the quality of the component mounting substrate due to the poor processing state of the processing apparatus.

The component mounting system may be configured as follows: the third determination unit determines whether the processing status indicated by the first determination information is good or not, and the processing status indicated by the second determination information is good or not. When the results do not match, the third judgment information indicating that the final judgment result is that the processing state is good is output.

In this aspect, when the judgment results indicated by the first judgment information and the second judgment information do not match, the third judgment unit outputs the third judgment information of the processing device indicating that the final judgment result is that the processing state is good. In this case, it is possible to prevent the determination criteria of the final determination by the third determination unit from being set too strict.

As described above, according to the present invention, the determination accuracy of the determination based on the image related to the processing state of the processing device in the component mounting line can be improved.

2: Install the body of the machine 3: Install the camera 4: Install the control unit 10: Parts installation line 11: Processing device 11A: Solder printing device 11B: Parts Mounting Device 11C: Reflow device 12: Inspection device 12A: Solder Paste Inspection Device 12B: Equipped with inspection device 12C: Reflow inspection device 13: Management data memory device 14: Algorithm memory device 15: Management device 21: Ontology framework 23: Conveyor 24: Parts supply unit 24F: Feeder 25: Head unit 27: Mobile Frames 28: Substrate support unit 31: The first camera department 32: The second camera department 33: The third camera department 40: Install the Communications Department 41: Communication Control Department 42: Board conveyance control section 43: Parts Supply Control Department 44: Head Control Department 45: Camera Control Department 46: The first adsorption determination part 47: Second Adsorption Determination Section 48: The third adsorption determination part 100: Parts Mounting System 121: Check Communications Department 122: Check the camera department 123: Check Control Department 151: Manage Communications Department 152: Display part 153: Operation Department 154: Management Control Department 251: carrying head 261: Fixed track 262: Ball screw shaft 263: Y-axis servo motor 271: Nut part 272: Guide member 273: Ball screw shaft 274: X-axis servo motor 1231: Communication Control Department 1232: Camera Control Department 1233: The 1st Inspection and Judgment Section 1234: Second Inspection and Judgment Section 1235: The 3rd Inspection and Judgment Section 1541: Communication Control Department 1542: Display Control Department 1543: Category Information Generation Department 1544: Learning Materials Generation Department 1545: Learning Department 2511: adsorption nozzle PP: printed solder paste substrate PPA: Parts Mounting Substrate PPB: reflow substrate

FIG. 1 is a diagram showing the overall configuration of a component mounting system according to an embodiment of the present invention. FIG. 2 is a block diagram of a component mounting device included in the component mounting system. Fig. 3 is a plan view showing the structure of a mounting machine body of the component mounting device. Fig. 4 is an enlarged view showing the head unit portion of the mounting machine body. FIG. 5 is a flowchart showing the parts mounting operation of the parts mounting apparatus. FIG. 6 is a block diagram of an inspection device included in the component mounting system. FIG. 7 is a flowchart showing the inspection determination operation of the inspection apparatus. FIG. 8 is a diagram showing management data stored and memorized in a management data storage device included in the component mounting system. FIG. 9 is a block diagram of a management device included in the component mounting system. FIG. 10 is a flowchart showing the flow of learning processing in the management device.

12: Inspection device

121: Check Communications Department

122: Check the camera department

123: Check Control Department

1231: Communication Control Department

1232: Camera Control Department

1233: The 1st Inspection and Judgment Section

1234: Second Inspection and Judgment Section

1235: The 3rd Inspection and Judgment Section

Claims (7)

A part mounting system having: A component mounting line comprising: a plurality of processing devices including a component mounting device for producing a component-mounted substrate for mounting components, and one or a plurality of inspection devices for inspecting the processing status of the above-mentioned processing devices; and A management device that manages the above-mentioned part installation line; and At least any plurality of specific devices in the above-mentioned component mounting device and the above-mentioned inspection device respectively include: an imaging unit, which acquires images of the processing state of the processing device; a first judging unit for judging whether the processing state is good or not by detecting the feature quantity of the image, and outputting first judging information indicating the judgment result; a second determination unit that classifies the image into a first category and a second category using an image classification method based on artificial intelligence, and determines that the processing state is good when the image belongs to the first category, When the above-mentioned image belongs to the above-mentioned second category, determine that the above-mentioned processing state is bad, and output the second determination information indicating the result of the determination; and a third determination unit that outputs, based on the first determination information and the second determination information, third determination information indicating a final determination result of whether the processing state is good or not; and The above management device includes: an operation unit for inputting type designation instruction information, the type designation instruction information designating which of the above-mentioned first type and the above-mentioned second type the images corresponding to the plurality of specific devices belong to; A class information generation unit that generates class information based on the third determination information and the class designation command information corresponding to the most downstream device arranged in the most downstream of the component mounting line among the plurality of specific devices, and the class information indicates the above Which of the above-mentioned first category and the above-mentioned second category the above-mentioned images corresponding to the plurality of specific devices belong to; a learning data generating unit that generates learning data that associates the image and the category information in correspondence with the second determination unit of each of the plurality of specific devices; and A learning unit for learning a classification criterion when the second determination part of the plurality of specific devices classifies the image based on the learning data. The component mounting system according to claim 1, wherein the category information generating unit generates information indicating that the plurality of specific devices are respectively determined when the third determination information corresponding to the most downstream device is information indicating a determination result that the processing state is good. Information that the corresponding image above belongs to the above-mentioned first category is regarded as the above-mentioned category information. The component mounting system of claim 1, wherein the third determination information corresponding to the most downstream device by the category information generating unit is information indicating a determination result that the processing state is good, and the most downstream device among the plurality of specific devices When the third determination information corresponding to the other specific devices other than the device is information indicating the determination result of the processing state being bad, it is generated indicating that the images corresponding to the plurality of specific devices belong to the category specified by the category specifying instruction information. The information of the category is regarded as the above-mentioned category information. The component mounting system according to claim 1, wherein the type information generating unit generates information indicating that the plurality of specific devices are respectively determined when the third determination information corresponding to the most downstream device is information indicating the determination result of the processing state being poor. The information corresponding to the above-mentioned image belonging to the category specified by the above-mentioned category specifying instruction information is regarded as the above-mentioned category information. The component mounting system according to claim 1, wherein the determination result of the processing state indicated by the first determination information by the third determination unit is consistent with the determination result of the processing status indicated by the second determination information In this case, the above-mentioned third judgment information indicating the matching judgment result as the final judgment result is output. The component mounting system of claim 1, wherein the third determination unit does not match the determination result of whether the processing state indicated by the first determination information is good and the determination result of whether the processing state indicated by the second determination information is good In this case, the third judgment information indicating that the final judgment result is that the processing state is not good is output. The component mounting system of claim 1, wherein the third determination unit does not match the determination result of whether the processing state indicated by the first determination information is good and the determination result of whether the processing state indicated by the second determination information is good In this case, the third determination information indicating that the final determination result is that the processing state is good is output.

Images